Authors: Linda Lee, Joong Hwa Lee, Serge Egelman, David Wagner
This is a paper for WEIS 2015.
- Submission (full papers) 27 February 2015
- Acceptance notification 10 April 2015
- Final papers (revisions) 15 May 2015
- Conference dates 22-23 June 2015
=============================
address why not Cochnan’s Q test or Anova and add a footnote to justify why not.clearer demographics and compare our demographic w/ general population- present more about the regression model. (equations with constants and whatnot)
- Develop more of the nonlinearity—look into interactions within data and recipient (like at end of section 3.1.2, less ad-hoc and for all the questions). without that, don’t say things like: "Regardless of the data recipient..." (3.1.1).
Before anything else:
- ~~better intro and framing (AND SEND TO SERGE AND DAVE)!
- ~~heavy editing pass again to format the paper well
- BETTER RELATED WORK; redo search for papers (human attention; econ-y things)
The rest:
- discussion on “middle” questions and question distributions
- beef up discussion of tech ratings with new risk/ben tables
- analysis and discussion on high-variance (table with data and recipient sorted by standard deviation for our own looking, then just talk about them in the paper without including them inside)
=============================
- use the full scale instead of 5/not for regression model (reviewer suggestion)
- "more discussion," "more tables" .. there was no room.
- reviewer said: “3. The authors make heavy use of randomizing the order of measurement. Whenever this method is employed, one should make sure that the order indeed did not affect the results.”